Project Summary Inflammatory arthritis in adults and children is often characterized by periods of quiescent activity followed by disease flares. In any individual patient, the same joints typically flare repeatedly, in a pattern that usually establishes itself early in disease and then persists for years or decades. The hypothesis of this proposal is that this ?joint-specific memory? reflects the presence of T resident memory (TRM) cells, whose targeting represents a new approach to arthritis therapy. TRM are a recently-described subset of long-lived T cells, either CD8 and CD4, that develop in skin and other tissues as a response to tissue inflammation, persisting for years thereafter to provide long-lasting site-specific immunity. TRM have never been described in joints. We have now developed compelling evidence for the presence of these cells in human arthritic synovium using three orthogonal approaches: single-cell RNAseq, a novel 3-dimensional synovial culture system, coupled with cytometry by time of flight (CyTOF), and Mantra multidimensional immunofluorescence imaging. Further, we have developed or adapted three animal models to develop new murine systems optimized for the study of recurrent, joint-specific, T cell-dependent inflammatory arthritis. Building upon these preliminary data, we propose two specific aims. First, we will perform a comprehensive characterization of human synovial TRM with respect to surface phenotype, mediators, transcriptome, and metabolism to compare synovial TRM with synovial effector memory T cells and ?gold standard? TRM from human skin. Second, we will use our animal models to characterize the development and persistence of TRM over time; to test the possibility that these cells can re-activate arthritis upon antigen-independent triggering as well as antigen exposure; and employ local depletion to test TRM targeting as a novel approach to durable joint-specific arthritis therapy.